Measuring Spatial Access to Resources at Scale

Who am I

• Graduated from Brophy College Preparatory in Phoenix, AZ
• Studied a mix of Geography, Public Policy, and Computer Science
• Data scientist at the Urban Institute:
  • Build tools to support communities’ support fair resource allocation and upward mobility from poverty

Run of Show

• Motivation: Why is spatial access important
• Present on some spatial and demographic data analysis concepts
• Break you all in three groups to have you all consider a spatial access measure
• We will share out to the group
• Real life example
• I will present a bit on how we think about these things in practice for a piece of software I have co-built and work on at Urban

Motivation

• Government resource allocation: Identify disparities or suitable opportunities for investment
• Business (location intelligence): Determine underserved markets
• Assess property values: Proximity to amenities affects real estate pricing

Key Concepts: Point and Polygon Data

Point Data

Polygon Data

Census Tracts

Definition: Census tracts are small, relatively permanent statistical subdivisions of a county defined by the U.S. Census Bureau. They typically contain between 1,200 and 8,000 people, with an optimal size of 4,000 people.

• Often closely aligned with a concept of a “neighborhood” in cities
• Rural areas: they are much larger (to encompass enough population)
• Boundaries generally follow visible features (roads, rivers) and governmental boundaries

Census Tracts: Urban vs. Rural

St. Alban’s Census Tract

Centroids

• Informally: The center of a polygon
• More formally: The geometric center or “average location” of all points in a polygon

Spatial Access Methods: Group Activity

• Break into three groups
• Each group reviews one method on the following slides
• Your tasks:
  • Review and understand the approach
  • Determine the strengths and limitations of the approach
  • Identify an example for when the method would be suitable and unsuitable
  • Nominate a group representative to share out your results

Access the Slides

https://bit.ly/albans_access

Method 1: The Container Method

Steps:

1. Count the number of points (resources) in a given census tract
2. Treat all points in the polygon as accessible to residents of the census tract
3. Treat all points outside the census tract as inaccessible
4. Optionally, divide number of points by the population of the tract to get a “resource per person” measure

Method 2: Minimum Distance

Steps:

1. For each tract, start with a tract centroid
2. Calculate the minimum distance from the centroid to a given resource
3. The closer the nearest resource, the higher the access

Method 3: Gravity Potential

Steps:

1. Start with the tract centroid
2. For the tract centroid:
   • Calculate the distance to each resource
   • Divide 1 by the square of each distance
   • Sum those values

Discussion Questions

For your assigned method, discuss:

• What are the strengths?
• What are the limitations?
• Provide a use case that demonstrates the strength
• Provide a use case where the measure is inappropriate

Example: DC Libraries

Spatial Equity Data Tool

• Free software that lives on Urban’s website
• Built to respond to local government desires to advance racial equity and align with “smart cities” goals
• Cities want analytics but may not have support to calculate it
• Or they have different definitions of access/equity
• Uses the container method

Demo

Live demonstration of the Spatial Equity Data Tool

Visit: https://apps.urban.org/features/equity-data-tool/

Thank You

1. Questions? Ask me anything!

2. Takeaway: When you travel from your home to some other place, think about the right way to measure how easy it is to get there